NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results

Sangmin Lee; Eunpil Park; Angel Canelo; Hyunhee Park; Youngjo Kim; Hyung Ju Chun; Xin Jin; Chongyi Li; Chun Le Guo; Radu Timofte; Qi Wu; Tianheng Qiu; Yuchun Dong; Shenglin Ding; Guanghua Pan; Weiyu Zhou; Tao Hu; Yixu Feng; Duwei Dai; Yu Cao; Peng Wu; Wei Dong; Yanning Zhang; Qingsen Yan; Simon J. Larsen; Ruixuan Jiang; Senyan Xu; Xingbo Wang; Xin Lu; Marcos V. Conde; Javier Abad-Hernandez; Alvaro Garcia-Lara; Daniel Feijoo; Alvaro Garcia; Zeyu Xiao; Zhuoyuan Li

doi:10.1109/CVPRW67362.2025.00099

NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results

Sangmin Lee
, Eunpil Park
, Angel Canelo
, Hyunhee Park
, Youngjo Kim
, Hyung Ju Chun
, Xin Jin
, Chongyi Li
, Chun Le Guo
, Radu Timofte
, Qi Wu
, Tianheng Qiu
, Yuchun Dong
, Shenglin Ding
, Guanghua Pan
, Weiyu Zhou
, Tao Hu
, Yixu Feng
, Duwei Dai
, Yu Cao

Peng Wu, Wei Dong, Yanning Zhang, Qingsen Yan, Simon J. Larsen, Ruixuan Jiang, Senyan Xu, Xingbo Wang, Xin Lu, Marcos V. Conde, Javier Abad-Hernandez, Alvaro Garcia-Lara, Daniel Feijoo, Alvaro Garcia, Zeyu Xiao, Zhuoyuan Li

School of Computer Science

Samsung
Nankai University
University of Würzburg
Imvision
Northwestern Polytechnical University Xian
The Second Affiliated Hospital of Xi’an Jiaotong University
Xi an Institute of Optics and Precision Mechanics of Cas
Xi'an University of Architecture and Technology
Esoft Systems A/S
University of Science and Technology of China
CidautAI
National University of Singapore

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

28 Scopus citations

Abstract

This paper reviews the NTIRE 2025 Efficient Burst HDR and Restoration Challenge, which aims to advance efficient multi-frame high dynamic range (HDR) and restoration techniques. The challenge is based on a novel RAW multi-frame fusion dataset, comprising nine noisy and misaligned RAW frames with various exposure levels per scene. Participants were tasked with developing solutions capable of effectively fusing these frames while adhering to strict efficiency constraints: fewer than 30 million model parameters and a computational budget under 4.0 trillion FLOPs. A total of 217 participants registered, with six teams finally submitting valid solutions. The top-performing approach achieved a PSNR of 43.22 dB, showcasing the potential of novel methods in this domain. This paper provides a comprehensive overview of the challenge, compares the proposed solutions, and serves as a valuable reference for researchers and practitioners in efficient burst HDR and restoration.

Original language	English
Title of host publication	Proceedings - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025
Publisher	IEEE Computer Society
Pages	993-1008
Number of pages	16
ISBN (Electronic)	9798331599942
DOIs	https://doi.org/10.1109/CVPRW67362.2025.00099
State	Published - 2025
Event	2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025 - Nashville, United States Duration: 11 Jun 2025 → 12 Jun 2025

Publication series

Name	IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops
ISSN (Print)	2160-7508
ISSN (Electronic)	2160-7516

Conference

Conference	2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025
Country/Territory	United States
City	Nashville
Period	11/06/25 → 12/06/25

Keywords

efficient burst hdr
restoration

Access to Document

10.1109/CVPRW67362.2025.00099

Cite this

Lee, S., Park, E., Canelo, A., Park, H., Kim, Y., Chun, H. J., Jin, X., Li, C., Guo, C. L., Timofte, R., Wu, Q., Qiu, T., Dong, Y., Ding, S., Pan, G., Zhou, W., Hu, T., Feng, Y., Dai, D., ... Li, Z. (2025). NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results. In Proceedings - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025 (pp. 993-1008). (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops). IEEE Computer Society. https://doi.org/10.1109/CVPRW67362.2025.00099

Lee, Sangmin ; Park, Eunpil ; Canelo, Angel et al. / NTIRE 2025 Challenge on Efficient Burst HDR and Restoration : Datasets, Methods, and Results. Proceedings - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025. IEEE Computer Society, 2025. pp. 993-1008 (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops).

@inproceedings{c8a11f22c2224a2e91644d935ebd6d9b,

title = "NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results",

abstract = "This paper reviews the NTIRE 2025 Efficient Burst HDR and Restoration Challenge, which aims to advance efficient multi-frame high dynamic range (HDR) and restoration techniques. The challenge is based on a novel RAW multi-frame fusion dataset, comprising nine noisy and misaligned RAW frames with various exposure levels per scene. Participants were tasked with developing solutions capable of effectively fusing these frames while adhering to strict efficiency constraints: fewer than 30 million model parameters and a computational budget under 4.0 trillion FLOPs. A total of 217 participants registered, with six teams finally submitting valid solutions. The top-performing approach achieved a PSNR of 43.22 dB, showcasing the potential of novel methods in this domain. This paper provides a comprehensive overview of the challenge, compares the proposed solutions, and serves as a valuable reference for researchers and practitioners in efficient burst HDR and restoration.",

keywords = "efficient burst hdr, restoration",

author = "Sangmin Lee and Eunpil Park and Angel Canelo and Hyunhee Park and Youngjo Kim and Chun, \{Hyung Ju\} and Xin Jin and Chongyi Li and Guo, \{Chun Le\} and Radu Timofte and Qi Wu and Tianheng Qiu and Yuchun Dong and Shenglin Ding and Guanghua Pan and Weiyu Zhou and Tao Hu and Yixu Feng and Duwei Dai and Yu Cao and Peng Wu and Wei Dong and Yanning Zhang and Qingsen Yan and Larsen, \{Simon J.\} and Ruixuan Jiang and Senyan Xu and Xingbo Wang and Xin Lu and Conde, \{Marcos V.\} and Javier Abad-Hernandez and Alvaro Garcia-Lara and Daniel Feijoo and Alvaro Garcia and Zeyu Xiao and Zhuoyuan Li",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025 ; Conference date: 11-06-2025 Through 12-06-2025",

year = "2025",

doi = "10.1109/CVPRW67362.2025.00099",

language = "英语",

series = "IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops",

publisher = "IEEE Computer Society",

pages = "993--1008",

booktitle = "Proceedings - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025",

}

Lee, S, Park, E, Canelo, A, Park, H, Kim, Y, Chun, HJ, Jin, X, Li, C, Guo, CL, Timofte, R, Wu, Q, Qiu, T, Dong, Y, Ding, S, Pan, G, Zhou, W, Hu, T, Feng, Y, Dai, D, Cao, Y, Wu, P, Dong, W, Zhang, Y , Yan, Q, Larsen, SJ, Jiang, R, Xu, S, Wang, X, Lu, X, Conde, MV, Abad-Hernandez, J, Garcia-Lara, A, Feijoo, D, Garcia, A, Xiao, Z & Li, Z 2025, NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results. in Proceedings - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025. IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops, IEEE Computer Society, pp. 993-1008, 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025, Nashville, United States, 11/06/25. https://doi.org/10.1109/CVPRW67362.2025.00099

NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results. / Lee, Sangmin; Park, Eunpil; Canelo, Angel et al.
Proceedings - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025. IEEE Computer Society, 2025. p. 993-1008 (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - NTIRE 2025 Challenge on Efficient Burst HDR and Restoration

T2 - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025

AU - Lee, Sangmin

AU - Park, Eunpil

AU - Canelo, Angel

AU - Park, Hyunhee

AU - Kim, Youngjo

AU - Chun, Hyung Ju

AU - Jin, Xin

AU - Li, Chongyi

AU - Guo, Chun Le

AU - Timofte, Radu

AU - Wu, Qi

AU - Qiu, Tianheng

AU - Dong, Yuchun

AU - Ding, Shenglin

AU - Pan, Guanghua

AU - Zhou, Weiyu

AU - Hu, Tao

AU - Feng, Yixu

AU - Dai, Duwei

AU - Cao, Yu

AU - Wu, Peng

AU - Dong, Wei

AU - Zhang, Yanning

AU - Yan, Qingsen

AU - Larsen, Simon J.

AU - Jiang, Ruixuan

AU - Xu, Senyan

AU - Wang, Xingbo

AU - Lu, Xin

AU - Conde, Marcos V.

AU - Abad-Hernandez, Javier

AU - Garcia-Lara, Alvaro

AU - Feijoo, Daniel

AU - Garcia, Alvaro

AU - Xiao, Zeyu

AU - Li, Zhuoyuan

PY - 2025

Y1 - 2025

N2 - This paper reviews the NTIRE 2025 Efficient Burst HDR and Restoration Challenge, which aims to advance efficient multi-frame high dynamic range (HDR) and restoration techniques. The challenge is based on a novel RAW multi-frame fusion dataset, comprising nine noisy and misaligned RAW frames with various exposure levels per scene. Participants were tasked with developing solutions capable of effectively fusing these frames while adhering to strict efficiency constraints: fewer than 30 million model parameters and a computational budget under 4.0 trillion FLOPs. A total of 217 participants registered, with six teams finally submitting valid solutions. The top-performing approach achieved a PSNR of 43.22 dB, showcasing the potential of novel methods in this domain. This paper provides a comprehensive overview of the challenge, compares the proposed solutions, and serves as a valuable reference for researchers and practitioners in efficient burst HDR and restoration.

AB - This paper reviews the NTIRE 2025 Efficient Burst HDR and Restoration Challenge, which aims to advance efficient multi-frame high dynamic range (HDR) and restoration techniques. The challenge is based on a novel RAW multi-frame fusion dataset, comprising nine noisy and misaligned RAW frames with various exposure levels per scene. Participants were tasked with developing solutions capable of effectively fusing these frames while adhering to strict efficiency constraints: fewer than 30 million model parameters and a computational budget under 4.0 trillion FLOPs. A total of 217 participants registered, with six teams finally submitting valid solutions. The top-performing approach achieved a PSNR of 43.22 dB, showcasing the potential of novel methods in this domain. This paper provides a comprehensive overview of the challenge, compares the proposed solutions, and serves as a valuable reference for researchers and practitioners in efficient burst HDR and restoration.

KW - efficient burst hdr

KW - restoration

UR - https://www.scopus.com/pages/publications/105017858560

U2 - 10.1109/CVPRW67362.2025.00099

DO - 10.1109/CVPRW67362.2025.00099

M3 - 会议稿件

AN - SCOPUS:105017858560

T3 - IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops

SP - 993

EP - 1008

BT - Proceedings - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025

PB - IEEE Computer Society

Y2 - 11 June 2025 through 12 June 2025

ER -

Lee S, Park E, Canelo A, Park H, Kim Y, Chun HJ et al. NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results. In Proceedings - 2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, CVPRW 2025. IEEE Computer Society. 2025. p. 993-1008. (IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops). doi: 10.1109/CVPRW67362.2025.00099

NTIRE 2025 Challenge on Efficient Burst HDR and Restoration: Datasets, Methods, and Results

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this